Subpial cortical demyelination induced by subarachnoid injection of tumour necrosis factor and interferon gamma in MOG immunized animals

Christopher Gardner, Deanna Taylor, Richard Reynolds. Department of Cellular and Molecular Neuroscience, Division of Neuroscience and Mental Health, Imperial College London, Hammersmith Hospital Campus, Burlington Danes Building, Du Cane Road, London W12 0NN, UK.

Recent studies of cerebral cortical pathology in secondary progressive multiple sclerosis (MS) have shown the presence of extensive sub-pial grey matter lesions in a significant proportion of cases. The association of sub-pial lesions with ectopic B-cell follicles in the cerebral meninges suggests that the diffusion of pro-inflammatory cytokines from the meninges into the brain parenchyma could be responsible for this pathology. To test this hypothesis we have established an animal model mimicking the cortical pathology observed in MS. Female DA rats were immunized with 5 micrograms recombinant myelin oligodendrocyte glycoprotein (MOG) in incomplete Freunds adjuvant (IFA). Although this dose was insufficient to cause clinical symptoms, a humoral (IgG) immune response was raised to MOG. Twenty days post immunization, animals received an injection of tumour necrosis factor (TNF) and interferon-gamma (IFNG) into the sub-arachnoid space at the sagittal sulcus.

Immunohistochemistry revealed areas of subpial demyelination extending from the sulcus over the dorsal surface of the cortex, encompassing layers I-III, associated with extensive activation of microglia/macrophages. Lesions were first evident at 3 days post injection, were maximal after 7 days and had resolved by remyelination at 14 days. The extent of demyelination correlated with the number of ED1+/Iba1+ microglia/macrophages. Cytotoxic CD8+ T cells were observed in the meninges, corpus callosum and scattered throughout the grey matter, whereas CD4+ T cells and CD79a+ B cells were restricted to the meninges. Control animals (IFA alone injected with cytokines) had increased numbers of meningeal macrophages and lymphocytes compared to naïve animals, but had little or no demyelination despite cortical microglial activation. Animals immunised with MOG and injected with phosphate buffered saline showed no increase in meningeal macrophages or lymphocytes and no cortical demyelination.

These findings are the first report of an animal model of sub-pial demyelination resulting from the presence of pro-inflammatory cytokines in the sub-arachnoid space and support our hypothesis of a role for meningeal inflammation in the cortical pathology of MS. The subpial demyelination was dependent on a pre-existing immune response against a myelin protein. The pattern of lymphocyte infiltration is similar to that seen in post-mortem human tissue, indicating that the model has potential as a model of MS cortical pathology.